Microbiome Composition Within a Sympatric Species Complex of Intertidal Isopods (Jaera Albifrons)

RESEARCH ARTICLE Microbiome composition within a sympatric species complex of intertidal isopods (Jaera albifrons)

Marius A. Wenzel*, Alex Douglas, Stuart B. Piertney

School of Biological Sciences, University of Aberdeen, Aberdeen, United Kingdom

* [email protected] a1111111111 a1111111111 a1111111111 Abstract a1111111111 a1111111111 The increasingly recognised effects of microbiomes on the eco-evolutionary dynamics of their hosts are promoting a view of the ªhologenomeº as an integral host-symbiont evolutionary entity. For example, sex-ratio distorting reproductive parasites such as Wolbachia are well-studied pivotal drivers of invertebrate reproductive processes, and more recent work is OPEN ACCESS highlighting novel effects of microbiome assemblages on host mating behaviour and devel- Citation: Wenzel MA, Douglas A, Piertney SB opmental incompatibilities that underpin or reinforce reproductive isolation processes. How- (2018) Microbiome composition within a ever, examining the hologenome and its eco-evolutionary effects in natural populations is sympatric species complex of intertidal isopods challenging because microbiome composition is considerably influenced by environmental (Jaera albifrons). PLoS ONE 13(8): e0202212. https://doi.org/10.1371/journal.pone.0202212 factors. Here we illustrate these challenges in a sympatric species complex of intertidal isopods (Jaera albifrons spp.) with pervasive sex-ratio distortion and ecological and beha- Editor: Cheryl S. Rosenfeld, University of Missouri Columbia, UNITED STATES vioural reproductive isolation mechanisms. We deep-sequence the bacterial 16S rRNA gene among males and females collected in spring and summer from two coasts in north- Received: February 20, 2018 east Scotland, and examine microbiome composition with a particular focus on reproductive Accepted: July 29, 2018 parasites. Microbiomes of all species were diverse (overall 3,317 unique sequences among Published: August 29, 2018 3.8 million reads) and comprised mainly Proteobacteria and Bacteroidetes taxa typical of Copyright: © 2018 Wenzel et al. This is an open the marine intertidal zone, in particular Vibrio spp. However, we found little evidence of the access article distributed under the terms of the reproductive parasites Wolbachia, Rickettsia, Spiroplasma and Cardinium, suggesting alter- Creative Commons Attribution License, which native causes of sex-ratio distortion. Notwithstanding, a significant proportion of the vari- permits unrestricted use, distribution, and reproduction in any medium, provided the original ance in microbiome composition among samples was explained by sex (14.1 %), nested author and source are credited. within geographic (26.9 %) and seasonal (39.6 %) variance components. The functional rel- Data Availability Statement: Raw sequence data evance of this sex signal was difficult to ascertain given the absence of reproductive para- alongside all relevant metadata are deposited in the sites, the ephemeral nature of the species assemblages and substantial environmental NCBI short-read archive (SRA) at accession variability. These results establish the Jaera albifrons species complex as an intriguing sys- SRP132549. tem for examining the effects of microbiomes on reproductive processes and speciation, Funding: This work was funded by the UK Natural and highlight the difficulties associated with snapshot assays of microbiome composition in Environment Research Council (NERC), grant NE/ M015661/1, awarded to SBP and AD. The funders dynamic and complex environments. had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 1 / 19 Microbiome composition in Jaera albifrons spp.

Introduction Microbiomes have long been recognised as important functional extensions of their host’s physiological and broader ecological phenotype. For example, microbiomes affect fundamental physiological processes associated with digestion, immune system function, disease aetiol- ogy and behaviour [1–3], ecological processes such as nutrient cycling at the plant-root/soil interface [4, 5] and calcification, proliferation and community structure of coral reefs [6, 7], as well as key evolutionary transitions such as the origin of mitochondria [8], gain of photosynthetic function in eukaryotic cells [9] or the parallel and convergent evolution of biolumines- cent “light organs” in squid and angler fishes [10, 11]. From an evolutionary perspective, microbiome composition is also implicated in reproductive isolation and speciation via affecting chemosensory cues essential for mating preference [12] or causing fundamental developmental incompatibilities and hybrid breakdown [13]. These insights have given rise to the “hologenome” concept of considering the host macro-organism and its associated microbiome as an integral evolutionary entity [14–16]. As such, studying the multi-layered effects of host- microbiome interactions holds immense value for a broad array of pure and applied disci- plines ranging from medicine and agriculture to molecular physiology and ecosystem ecology and evolution [14, 17, 18]. A centrally important phenomenon that underlines how microbiomes may affect eco-evolutionary processes in their hosts is sex-ratio distortion in invertebrates caused by infection with cytoplasmic reproductive endoparasites. Wolbachia, Rickettsia, Spiroplasma, Cardinium bacteria and Microsporidian fungi infect the reproductive organs of many arthropod and nematode species and are cytoplasmically transmitted from mother to offspring [19–23]. As a means of promoting transmission and infection prevalence in the population, these parasites manipulate host reproductive biology to distort host sex-ratios in favour of infected females by induction of parthenogenesis, feminization of male offspring, killing of male embryos, disruption of sex-chromosome inheritance, or cytoplasmic incompatibility between individuals with different infection statuses [19, 22, 24]. This demographic disruption can lead to erosion of genetic diversity and phylogenetic signal akin to a bottleneck or selective sweep since most of the population will eventually be descended from few infected matrilines [25, 26]. Con- versely, Wolbachia infection can also promote diversification via horizontal gene transfer to the host [22], and initiation or reinforcement of reproductive isolation and speciation through cytoplasmic incompatibility between populations with mixed infections [27, 28]. Not least, Wolbachia infection can perturb overall microbiome composition, often in sex-specific fashion with downstream physiological effects [29–32]. In concert, these factors firmly establish Wol- bachia and other reproductive parasites as pivotal agents in driving the evolution of many invertebrates. Beyond the obvious value in studying prominently important taxa such as Wolbachia and other reproductive parasites, key to gaining a proper understanding of the effects of the hologenome on any facet of eco-evolutionary dynamics is the capacity to examine microbiome- wide patterns of diversity in free-living non-model systems. The wide availability of high- throughput DNA sequencing has enabled rapid characterisation of microbial species composition in virtually any type of field sample, and is poised to revolutionise our understanding of how the hologenome operates and evolves in the wild [33, 34]. However, a pre-requisite to thoroughly understanding microbiome composition is an appreciation of potential environmental sources of variation. Abiotic factors in dynamic natural environments may confer considerable spatio-temporal variation in ephemeral uptake and proliferation of commensal microbionts that may not necessarily be functionally linked to host metabolism. For example, microbiome composition in marine copepods is contingent on seasonal and spatial differences

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 2 / 19 Microbiome composition in Jaera albifrons spp.

in water temperatures [35], and littoral Hymeniacidon heliophila sponges can display small- scale variation in microbiome composition between subtidal and intertidal specimens [36]. Conversely, the microbiomes of various marine nematode species are not obviously structured across habitats even on a global scale [37]. These examples highlight the need for an initial assessment of the degree of environmental variation in microbiomes before attempts are made to identify functionally relevant variation in the hologenome and its contribution to host ecology and evolution [38]. The Jaera albifrons (sensu lato) species complex of sympatric intertidal isopods is an intui- tively attractive study system for examining the links between reproductive parasites, microbiomes and host eco-evolutionary processes. In Europe, the complex comprises Jaera albifrons sensu stricto, Jaera ischiosetosa, Jaera praehirsuta and Jaera forsmani, with Jaera nordmanni as a congeneric outgroup taxon [39–41]. All species are common across North Atlantic coasts and often form mixed populations in sympatry or parapatry along the intertidal zone. The ingroup species are reproductively isolated through female preference for tactile courtship stimuli administered by males [41, 42], genetic incompatibilities conferring rapid hybrid breakdown [41, 42], and ecological zonation due to species-specific preferences of substrate, drainage, salinity and exposure [41, 43, 44], though some plasticity in the degree of reproductive isolation and frequency of introgressive hybridisation has been noted [41, 45]. In spite of these reproductive isolation mechanisms, all species are polyphyletic according to the mitochondrial 16S rRNA gene and also usually display substantial sex-ratio bias towards females [46–48]. This would be consistent with the presence of sex-ratio distorting reproductive parasites and associated erosion of mitochondrial diversity [42]. Attempts of detecting Wolbachia in Jaera via PCR have not yielded conclusive evidence for ongoing infection [48, 49]. However, a proper characterisation of the Jaera microbiome via next-generation sequencing has not yet been attempted, thus reproductive parasites other than Wolbachia may be present and affect Jaera demography and evolution. Moreover, such a characterisation would be an invaluable resource for exploring whether the Jaera microbiome could be involved in driving speciation and reproductive isolation mechanisms in the species complex, potentially through reinforcement of ecological niche partitioning via metabolic co-adaptation, affecting chemosensory or behavioural mate choice, or developmental hybrid incompatibility [12, 13, 50]. Here we use next-generation amplicon sequencing to provide a first characterisation of the microbiomes of males and females across the Jaera albifrons species complex. We deep- sequence the V3/V4 region of the bacterial 16S rRNA gene from DNA pools of Jaera individuals, examine specifically whether the reproductive parasites Wolbachia, Rickettsia, Spiroplasma and Cardinium are present, and explore broader signatures of seasonal, spatial and sex-specific variation in microbiome composition from samples collected in spring and summer from two coasts in north-east Scotland. This initial description of the Jaera microbiome will develop hypotheses for factors affecting microbiome composition in Jaera and establish the Jaera albifrons species complex as a powerful system for investigating the role of the microbiome in reproductive processes and speciation.

Materials and methods Sample collection and processing Jaera spp. are common intertidal invertebrates that are neither protected nor require sampling permits. Live individuals were collected in spring and summer 2017 from two coasts in north- east Scotland, separated by c. 200 km of coast line. Gardenstown on the north coast (57.672 ˚N, –2.337 ˚E) harbours all four European species of the species complex alongside the outgroup Jaera nordmanni in varying composition along the shoreline. Two beaches on the

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 3 / 19 Microbiome composition in Jaera albifrons spp.

south-east coast in close vicinity to each other (Johnshaven: 56.796 ˚N, –2.328 ˚E; Arbroath: 56.518 ˚N, –2.659 ˚E) harbour >95 % pure single-species populations of Jaera albifrons and Jaera ischiosetosa respectively. As soon as possible after collection, individuals were sexed and assigned to species by identifying patterns of pereiopod setation in males using light microscopy [40, 41]. Identified males were then kept at room temperature in a large tub containing filtered sea water from the collection site. Since females cannot be assigned to species morphologically [40, 41], we assigned putative species based on species composition of males collected at the same beach section. An approximately even species mix of females was added to the same tub as the males, and all individuals were starved for one week to reduce gut content. Individuals were then briefly rinsed in sterile water and immediately processed for DNA extraction. We generated eleven DNA samples that comprised one pure male pool (6-12 individuals) for each of all five species, five mixed or presumably pure female pools (6-12 individuals), and a single exceptionally large female of unknown species (Table 1). These samples not only allowed us to screen males and females of all species for sex-ratio distorting parasites, but also to capture a broad snapshot of the Jaera microbiome across coasts and seasons. Samples were homogenized in heated (60 ˚C) lysis buffer (0.1 M Tris-HCl, 0.05 M EDTA, 0.1 M NaCl, 1 % SDS, 400 μg Proteinase K, 100 μg RNAse A) using an autoclaved teflon plunger. The homogenate was incubated overnight at 60 ˚C and DNA was extracted via standard phenol- chloroform extraction. DNA quality and quantity were checked with a NanoDrop ND-1000 spectrophotometer. Samples were submitted to Eurofins Genomics (Ebersberg, Germany) for bacterial 16S rRNA V3/V4 amplicon generation (c. 420 bp) using the standard S-D-Bact- 0341-b-S-17 and S-D-Bact-0785-a-A-21 primers [51] with sample-specific barcodes, and sequencing on the Illumina MiSeq V3 platform in 300 bp paired-end mode.

Sequence assembly, curation and taxonomic classification Raw sequence reads were filtered, de-noised and assembled to unique single-end (forward reads only) as well as paired-end sequence variants using DADA2 v1.4.0 [52] in R v3.4.0 [53]. Reads were trimmed at nucleotide call quality below 2, and reads with undetermined bases were discarded. Exploratory quality plots indicated a rapid decline in basecall quality towards the ends of the reads, particularly for reverse reads. Therefore, for paired-end analysis, forward reads were further trimmed to 260 bp and reverse reads to 220 bp, ensuring an overlap of at least 60 bp. Error rates were estimated and sequences were de-noised separately for forward and reverse reads in pooled-sample mode. Single-end and paired-end contigs were assembled from de-noised data and chimera sequences were removed. Taxonomic classification to genus level was assigned from the SILVA NR v128 database [54] using the RDP classifier algorithm [55] with k-mer size 8, 100 bootstrap replicates and a minimum bootstrap support of 50. Species-level classification was added where possible based on 100 % sequence identity with SILVA NR v128 [56]. Sequences that were assigned to chloroplast, mitochondria, archaea, or eukaryota taxa were removed. Each sample was further annotated with putative functional metabolic capabilities of the identified microbial community using TAX4FUN v0.3.1 and associated pre-computed SILVA NR v123 reference data [57]. The observed sequence counts were transformed into abundances of KEGG enzymes via association with KEGG reference organisms. These enzymes were further classified with the first three levels in KEGG functional hierarchies [58]. The taxonomically classified microbiomes were then screened for reproductive parasite species in the Wolbachia, Rickettsia, Spiroplasma and Cardinium genera or relevant higher taxonomic levels. Candidate sequences were aligned with all available Wolbachia, Rickettsia,

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 4 / 19 Microbiome composition in Jaera albifrons spp.

Table 1. Summary of sequencing effort and sequence diversity across eleven samples.

ID Species Season Region Sex Single-end Paired-end Variants Chao1 ACE Shannon Simpson InvSimpson Fisher reads reads S1 Jaera Spring North M 355,151 323,324 1,919 2,011 ± 19.433 1,993 ± 21.732 4.938 0.973 37.572 266.718 forsmani S2 Jaera Spring North M 285,237 254,668 1,998 2,096 ± 19.66 2,078 ± 22.451 4.719 0.955 22.429 289.874 albifrons S3 Mix Spring North F 239,064 218,576 1,409 1,579 ± 40.387 1,544 ± 16.246 5.404 0.984 64.391 198.618 S4 Mix Spring South F 298,546 279,462 1,019 1,342 ± 75.903 1,195 ± 16.267 3.660 0.860 7.127 131.909 S5 Jaera Summer North M 123,413 75,377 1,344 1,494 ± 29.587 1,469 ± 18.691 5.346 0.985 67.930 210.864 nordmanni S6 Jaera Summer North M 146,701 95,980 1,344 1,525 ± 35.511 1,474 ± 18.62 4.703 0.946 18.509 204.322 ischiosetosa S7 Jaera Summer North M 64,878 29,535 1,506 1,603 ± 20.221 1,589 ± 19.683 5.661 0.987 77.825 275.532 praehirsuta S8 Mix Summer North F 537,362 340,550 1,596 1,659 ± 16.183 1,643 ± 19.196 4.716 0.973 36.655 202.424 S9 Unknown Summer North F 776,641 517,273 1,584 1,628 ± 13.248 1,616 ± 18.159 4.389 0.956 22.781 190.543 S10 Jaera Summer South F 480,152 325,520 1,030 1,333 ± 46.439 1,372 ± 19.789 3.957 0.934 15.263 124.762 ischiosetosa S11 Jaera Summer South F 559,443 376,747 1,201 1,374 ± 32.229 1,345 ± 18.143 2.896 0.835 6.054 145.490 albifrons Total 3,866,588 2,837,012 3,317 – – – – – – Correlation single/paired end (Pearson’s r) – – 0.952 0.938 0.902 0.992 0.999 0.976 0.964 Association with season (P-value) 0.497 0.874 0.240 0.273 0.406 0.766 0.956 0.898 0.515 Association with region (P-value) 0.410 0.438 0.001 0.003 0.005 0.024 0.084 0.004 0.000 Association with sex (P-value) 0.002 0.021 0.045 0.118 0.120 0.057 0.137 0.232 0.006

Sample descriptors (species, season, region and sex) are given alongside numbers of de-noised single-end and paired-end reads, and the following diversity indices based on single-end reads: numbers of unique sequence variants, Chao1 ± SE, ACE ± SE, Shannon, Simpson, inverse Simpson and Fisher index. Below, Pearson’s correlation coefficient (r; all P 0.001) between single-end and paired-end datasets, and associations of metrics with sample descriptors (two-tailed Welch’s t-test P-value) are presented. Significant P-values (P 0.05) are emboldened. https://doi.org/10.1371/journal.pone.0202212.t001

Spiroplasma and Cardinium reference sequences in SILVA NR v128 using MAFFT V7.305 [59] and clustered using neighbour-joining on Kimura-2-parameter phylogenetic distances in APE v4.1 [60]. Sequences that clustered closely with the SILVA reference sequences were more closely examined using NCBI MEGABLAST [61] against the non-redundant nucleotide collection (NT).

Microbiome sequence diversity and composition Sequence diversity analyses were carried out on single-end as well as paired-end datasets, using R and the package PHYLOSEQ v1.19.1 [62]. Sequencing depths per sample were examined for circumstantial associations with the categorical sample variables season (spring and summer), region (north: Gardenstown; south: Johnshaven/Arbroath) and sex using negative binomial generalized linear models (GLM) in the MASS package [63]. Rarefaction curves were obtained by computing the number of unique sequence variants in subsamples of increasing sizes in steps of 1,000 sequences without replacement [64]. Diversity indices (Chao1, ACE, Shannon, Simpson, inverse Simpson and Fisher) were computed for each sample and compared between samples grouped by season, region or sex using two-tailed Welch’s t-test. Consistency of all metrics between single-end and paired-end datasets was examined using Pearson’s correlation test.

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 5 / 19 Microbiome composition in Jaera albifrons spp.

Rarefied versions of the datasets were obtained by subsampling to the lowest sequencing depth across samples. Microbiome structure between samples was explored with Jaccard and Bray-Curtis dissimilarity indices and visualised in two-dimensional space using metric (Jaccard) or non-metric (Bray-Curtis) multidimensional scaling (MDS). Samples were then clustered hierarchically using Ward’s criterion on the dissimilarity matrix, and clusters were visualised as dendrograms using GGTREE v1.6.10 [65]. Sources of variation attributed to season, region and sex were explored with distance-based redundancy analysis and permutational multivariate analysis of variance (PERMANOVA) with 9,999 permutations [66]. Finally, to further explore sex-specific differences in microbiome composition, we fitted negative binomial GLMs in a differential gene expression framework that accounts for differences in library size and dispersion, as implemented in DESEQ2 [67]. Fold changes were calculated between sexes accounting for season and region as covariates. P-values were corrected for multiple testing using the false-discovery rate method [68], and sequences with significant fold changes (FDR 0.1) were identified.

Results Taxonomic composition and diversity De-noised single-end sequence data comprised 64,878–776,641 reads that collapsed to 1,019– 1,998 unique sequence variants per sample. Across all eleven samples, 3,317 unique sequence variants were observed, which were assigned to 25 phyla, 45 classes, 94 orders, 185 families and 445 genera, based on the SILVA NR v128 database. However, considerable fractions of these sequence variants could not be assigned beyond particular taxon levels at the 50 % bootstrap cut-off, i.e., 0.78 % for phylum, 3.33 % for class, 8.06 % for order, 16.21 % for family, 42.15 % for genus and 94.16 % for species levels. Paired-end data recovered less diversity, comprising 29,535–517,273 reads and 329–1,800 unique sequence variants, and captured less diversity with 3,283 unique sequence variants assigned to 23 phyla, 40 classes, 89 orders, 182 families and 426 genera. However, taxonomy assignment was slightly better compared to single-end data, with non-classification rates of 0.45 % for phylum, 1.58 % for class, 4.28 % for order, 10.70 % for family, 35.93 % for genus and 93.97 % for species. Rarefaction curves approached asymptotic stages for most samples, suggesting that the sequencing effort captured the major- ity of sequence diversity in both types of datasets (S1 Fig). The Proteobacteria and Bacteroidetes were the dominant phyla in all samples, accounting for 69.7–94.0 % and 5.7–28.2 % of sequences per sample, and the six most abundant phyla accounted for 99.5–99.9 % (Fig 1). The six most abundant orders accounted for 62.4–92.7 % and the six most abundant genera for 23.3–58.6 % of sequences, of which Vibrio dominated most samples with up to 46.1 %. Notwithstanding, microbial sequence diversity was high in all samples, with a Simpson index of 0.835–0.987 and Fisher index of 125–290 (Table 1; S2 Fig). All metrics were highly correlated between single-end and paired-end datasets (r = 0.902 − 0.999;p 0.001; Table 1). Diversity was similar among seasons, but signatures of region and, in particular, sex were apparent in many diversity metrics (S2 Fig). Rigorous statistical analysis beyond basic Welch’s t-tests was precluded by low sample size, but these tests supported a difference between sexes in particular (Table 1; S2 Fig), consistent with shallower rarefaction curves in females (S1 Fig). Prediction of broad-brush metabolic capacity of the identified microbial communities via sequence similarity of taxonomically classified sequences to KEGG model taxa recovered 278 KEGG pathways. The proportion of sequences that could not be mapped to KEGG organisms (FTU) ranged from 55.0 % to 96.8 % (median 89.2 %) per sample. The most abundant top- level category was 09100 Metabolism (median 57.3 % across samples), followed by 09130

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 6 / 19 Microbiome composition in Jaera albifrons spp.

Fig 1. Relative sequence abundances of the six most abundant phyla, orders and genera across eleven samples (S1-S11), organised by season and sex. https://doi.org/10.1371/journal.pone.0202212.g001

Environmental Information Processing (23.9 %). The most abundant pathways included 09131 Membrane transport, 09102 Energy metabolism and 09101 carbohydrate metabolism (S3 Fig).

Sex-ratio distorting reproductive parasites

No sequences were directly assigned to known SILVA strains of the reproductive parasite genera Wolbachia, Rickettsia, Spiroplasma and Cardinium. However, some sequences were assigned to relevant higher taxonomic ranks: one sequence to the family Anaplasmatacea, 16 sequences to Rickettsiacea, 36 sequences to Flammeovirgaceae and one sequence to the order Entomo- plasmatales. The counts of these sequences ranged from 0 to 5,583, representing relative abundances of at best 1.16 % (Fig 2; S1 Table). Of these 54 sequences, three clustered reasonably closely with the Rickettsia, Spiroplasma and Cardinium clades, but no sequence clustered closely with Wolbachia (Fig 2).MEGABLAST broadly supported these classifications, matching an uncultured Rickettsiaceae bacterium (accession JQ701668.1) at 90 % identity to the Rickettsia sequence, and an uncultured bacterium from the Cytophaga-Flavobacterium-Bacteroides group (DQ812543.1) at 98 % identity to the Cardinium sequence. However, the presumed Spir- oplasma sequence matched with 94 % identity an uncultured bacterium from the Mycoplasma- taceae family (EU646196.1), which is situated in a different order than Spiroplasma.

Sources of variation in microbiome composition We further explored microbiome composition for sex-specific, seasonal and regional signatures. Since sequencing depth was different between male and female samples (negative

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 7 / 19 Microbiome composition in Jaera albifrons spp.

Fig 2. Neighbour-joining dendrogram (K2P phylogenetic distance) of SILVA NR v128 16S rRNA gene reference sequences for reproductive parasites Wolbachia, Rickettsia, Spiroplasma and Cardinium genera (dashed branches) and most closely related Jaera 16S rRNA gene sequences (families Rickettsiaceae and Flammeovirgaceae, and order Entomoplasmatales; solid branches). The relative sequence abundances of the Jaera sequences are summarised alongside. Tip labels correspond to sequence identifiers in S1 Table. https://doi.org/10.1371/journal.pone.0202212.g002

binomial GLM: z = 3.115;P = 0.002), the datasets were rarefied (64,878 sequences for single- end data and 29,535 sequences for paired-end data) to avoid spurious sex-specific signatures in microbiome composition. Ordination and hierarchical clustering of Jaccard dissimilarity among samples suggested two major clusters that correspond to samples collected in spring and summer. Within both seasons, samples are further clustered by geographic region, and the northern region is further subdivided by sex (Fig 3). Distance-based redundancy analysis ascribed 39.6 % of the total variance to season, 26.9 % to region and 14.1 % to sex, and all three hierarchical variance components were statistically significant (Table 2). Bray-Curtis dissimilarity broadly supported these patterns (Table 2), but hierarchical clustering did not consistently recover the same nested structure (S4 Fig). Since all samples from the southern regions were females, it cannot be ruled out that variance ascribed to region is in fact sex-specific variation. However, this is quite unlikely since variance among males and females in the northern region in summer was considerably smaller than the putative variance among regions in summer (Fig 3). Similarly, although not all species are represented within each cluster, it appears that structure among sexes outweighs structure associated with species or microgeography at the same beach. Jaera nordmanni and Jaera ischiosetosa males collected from the same set of rocks in Gardenstown (north) in summer

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 8 / 19 Microbiome composition in Jaera albifrons spp.

Fig 3. Hierarchical clustering (left) and metric multidimensional scaling (right) of Jaccard dissimilarity among samples. Sample categories (season, region and sex) are indicated by line type, symbol shape and colour, respectively. https://doi.org/10.1371/journal.pone.0202212.g003

clustered most closely, followed by Jaera praehirsuta males collected further down the shore at the same time. However, the single large female (unknown species) collected at the same beach and time did not cluster as closely with any of these three male single-species samples. Instead, she clustered most closely with a mix of females collected from the same rocks as the three male samples (Fig 3). Exploring the identified sex-specific signal in microbiome composition further with negative binomial models indicated that eleven sequences were significantly (FDR 0.1) more abundant in males and six sequences were more abundant in females, after accounting for differences in season and region (Fig 4). Of these 17 sequences ten had taxonomic annotation, representing ten genera in nine families: Aureispira, Peredibacter, Loktanella, Winogradskyella and Pibocella genera were more abundant in males, and Tenacibaculum, Marinomonas, Aliiro- seovarius, Leisingera and Pelagibius genera were more abundant in females (Fig 4). A simpli- fied analysis within the northern region only recovered similar patterns, corroborating

Table 2. Permutational multivariate analysis of variance (PERMANOVA) in Jaccard and Bray-Curtis dissimilarity indices among samples.

DF SS MS F R2 P Jaccard dissimilarity Season 1 0.931 0.931 10.141 0.396 0.000 Region 2 0.633 0.316 3.445 0.269 0.002 Sex 2 0.331 0.166 1.803 0.141 0.049 Residuals 5 0.459 0.092 0.195 – – Total 10 2.354 1.000 – – – Bray-Curtis dissimilarity Season 1 0.704 0.704 4.388 0.245 0.000 Region 2 0.772 0.386 2.407 0.268 0.001 Sex 2 0.599 0.299 1.866 0.208 0.018 Residuals 5 0.802 0.160 0.279 – – Total 10 2.876 1.000 – – –

Total variance was decomposed into hierarchical levels corresponding to season, region and sex, and statistical significance was estimated from 9,999 permutations. The table presents degrees of freedom (DF), sums of squares (SS), mean squares (MS), F-statistic, R-squared and P-value. Significant P-values (P 0.05) are emboldened.

https://doi.org/10.1371/journal.pone.0202212.t002

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 9 / 19 Microbiome composition in Jaera albifrons spp.

Fig 4. Sequence variants with differential abundance between sexes. The left panel summarises fold change and statistical significance for each sequence variant. The following two panels illustrate total aggregated sequence counts (abundance) and taxonomic classification (family or genus) of statistically significant (FDR 0.1) sequence variants. https://doi.org/10.1371/journal.pone.0202212.g004

Pibocella as male-associated and suggesting Vibrio and Owenweeksia as additional female-associated taxa (S5 Fig). Likewise, using non-rarefied data and a stricter significance threshold, a similar set of differentially abundant taxa was identified, highlighting Aliivibrio, Flavirhabdus and Polaribacter as further female-associated taxa (S6 Fig).

Discussion We present an initial survey of microbiome composition among males and females of all UK members of the Jaera albifrons species complex of intertidal isopods. The salient features of all microbiomes are high species diversity and absence of the classic feminizing reproductive parasite Wolbachia, though potentially novel strains of Rickettsia and Cardinium may be present instead. Additionally, microbiome composition varied considerably among samples and revealed hierarchical structure associated with season, region and sex. These patterns provide a first look at environmental sources of variation in microbial assemblages and could indicate an involvement of the microbiome in reproductive processes in Jaera.

Characterisation of Jaera microbiomes The microbial communities of all samples were dominated by the Proteobacteria and Bacteroi- detes phyla, which are widely described as the most abundant phyla in intertidal and open oce- anic environments [36, 69–72]. The high abundance of Vibrio in particular is consistent with microbiomes of other marine invertebrates such as copepods [35] or sea urchins [73]. Vibrio is a very common endo- and epibiont in marine crustaceans and often produces chitinolytic enzymes that allow for exploiting chitinous exoskeleton as a niche for attachment and proliferation [71, 74, 75]. Some Vibrio species are pathogens and others have been implicated in biogeochemical processes, but the specific metabolic relationships between crustaceans and Vibrio are cryptic [74]. Beyond these dominant taxa, sequence diversity in Jaera was high, consistent with diversity in intertidal sponges [36] and other marine invertebrates [35], and exceeded diversity of the terrestrial isopod Armadillidium vulgare [31]. A large proportion of sequences could not be taxonomically characterised to species level, and functional classification was hampered by very low mapping rate of sequences to KEGG organisms. Rarefaction curves suggested that more sequencing effort would have detected even more diversity in most samples, particularly

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 10 / 19 Microbiome composition in Jaera albifrons spp.

in males. This would suggest a wealth of uncharacterised taxonomic diversity, consistent with other studies investigating marine microbiomes [69, 76], and highlights the need for better reference characterisation of marine microbial communities [34]. In spite of capturing high microbial diversity, there was no evidence of known SILVA-curated strains of the reproductive parasites Wolbachia, Rickettsia, Spiroplasma and Cardinium. Although some sequences were classified to relevant higher taxonomic ranks and clustered rel- atively closely with known SILVA strains, only one sequence (presumed Rickettsia) formed a monophyletic group with known strains. Even if some of the identified sequences represented novel, somewhat diverged, strains of these reproductive parasites, all sequences in question had very low abundances and would not suggest high infection intensities. These results are difficult to reconcile with prevalent sex-ratio distortion in Jaera [46–48] and pervasive Wolba- chia infection in many crustaceans [49, 77]. However, these results are fully compatible with previous studies that have failed to reliably detect Wolbachia infection in total Jaera DNA extracts using targeted PCR assays [48, 49]. Ribardière et al. [48] screened 817 individuals across the Jaera albifrons species complex using 11 PCR protocols, but found little evidence of infection beyond an ephemeral novel haplotype in some Jaera albifrons and Jaera praehirsuta individuals, identified using a nested PCR protocol. As such, infection of Jaera species by Wol- bachia or other bacterial sex-ratio distorting parasites cannot be ruled out, but infection intensities and prevalence appear to be very low and difficult to detect. The biological relevance of rare sequences is difficult to assess and may well be an artefact of working with whole-body DNA extracts. Reproductive parasites primarily infect the reproductive and digestive tracts, thus it may be possible that dissection of these tissues prior to DNA extraction and sequencing improves detection [31, 78, 79]. Nevertheless, whole-body extracts should not in principle preclude detection of Wolbachia infection [29, 80], and even low infection levels should be readily detectable [81]. We thus conclude that prokaryotic reproductive parasites are unlikely to explain pervasive sex-ratio biases among the Jaera albifrons species complex, but note that the role of eukaryotic sex-ratio distorters (such as Microspori- dian fungi [23]) remains uninvestigated in Jaera.

Sources of variation in microbiome composition Microbiome composition in Jaera varied considerably between spring and summer despite similar species richness, suggesting that the microbiome in Jaera undergoes extensive temporal changes in species composition, yet remains fairly consistent in complexity. Seasonal changes in marine microbiomes are well-documented and are primarily driven by abiotic environmental factors such as temperature and biogeochemical processes [69, 82, 83]. In copepods, such seasonal changes have been reported even across a few weeks in early summer and may be linked to a rise in water temperatures and concomitant changes in temperature-sensitive marine microbial communities [35]. Any vacated ecological niches would then be taken up by different microbe species such that the overall species richness would not be greatly affected. For example, Vibrio form the core microbiome of copepods in subtropical locations, but Vibrio abundance is lower in temperate regions where a similar chitinolytic niche could be taken up by Pseudoalteromonas species [35]. The Jaera microbiome also showed regional structure in species richness and species composition across the two sampling regions in north-east Scotland. Although we sampled only females from the two southern coasts and therefore cannot rule out that at least part of the regional variance is attributable to species rather than geography, spatial effects on littoral and intertidal microbiome composition are, in fact, commonly reported at large and small scales. This is illustrated by vastly different microbiomes in Hymeniacidon heliophila sponges in

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 11 / 19 Microbiome composition in Jaera albifrons spp.

subtidal and intertidal habitats at the same site [36], as well as spatial structure in microbiomes in populations of sand hoppers across the Tuscan coast [70] and benthic amphipods in the Great Lakes [84]. The regional structure in the Jaera microbiome could be due to differences in abiotic factors such as aspect (north-facing versus east-facing beaches), geology (cliff coast vs. plain coast) and human activity (proximity to harbours or sewage outlets). One particular factor could be differences in salinity, either intrinsic to the substrate or extrinsic from nearby- freshwater streams. Both beaches in the south were affected by freshwater drainage from streams, whereas the beach in the north had no freshwater influx. The observed lower microbial diversity in the freshwater-affected sites contradicts the pattern found in a freshwater- marine transect in Greece [76], but is consistent with an increase in species richness and considerable changes in composition of the skin microbiome of Atlantic salmon after transition- ing from freshwater to seawater [85]. Finally, overall microbiome composition showed a signature of sex, nested within the larger environmental variance components. Males tended to have more diverse microbiomes than females and the sexes also differed in the presence and abundance of a range of taxa, including a range of putative pathogens such as Aliivibrio, Aliiroseovarius, Vibrio and Tenacibaculum [86–88], and common environmental species typical of marine arthropods with no immediate functional link to reproductive processes, such as Loktanella, Glaciecola, Aureispira, Wino- gradskyella, Pelagibius and Marinomonas [75, 89]. An interesting finding was the high abundance of Leisingera in females in summer, alongside Vibrio and Tenacibaculum. Secondary metabolites produced by Leisingera are known to have antimicrobial effects and are used by cephalopods to protect their eggs against pathogens such as Vibrio [90]. Nevertheless, the specific functional roles of the taxa assemblage in male and female Jaera remains obscure and will require more detailed functional assays and experimentation. Sex-specific differences in diversity and composition have been reported, for example, in whole-body microbiomes of phloem-feeding whiteflies, aphids and psyllids [29], and cloacal microbiomes of the striped plateau lizard Sceloporus virgatus [91]. In arthropods, these differences are often attributed to infection with reproductive endoparasites [29, 32]. For example, uninfected males and females of the terrestrial isopod Armadillidium vulgare have similar microbiomes, but Wolbachia-infected females carry higher total bacterial loads [30, 31]. Since we found little evidence of reproductive parasites in Jaera, alternative explanations need to be considered. One hypothesis could be that sex-specific microbiomes are functionally linked to intra-specific reproductive processes such as sex recognition or sexual selection that may have knock-on effects on reproductive isolation and speciation [12, 13, 92]. For example, reproductive isolation via pheromones is documented in allopatric populations of the marine polychaete Neanthes acuminata [93] and sympatric populations of the amphipod Eogammarus confervicolus [50]. The idea that microbiomes may be linked to host developmental processes and co-diverge tightly with host speciation events—a phenomenon termed “phylosymbiosis”—is a hotly debated topic [14–16, 38]. Although we were unable to separate species-specific signals from sex-specific signals with the present set of samples, the Jaera albifrons species complex would be an excellent study system for testing these ideas with more extensive sampling across both sexes within all species. Conversely, instead of enhancing metabolic function or causing behavioural changes in the host, sex-specific patterns in microbiont abundance could simply be attributed to differences in host body size or behaviour that circumstantially cause differential uptake and proliferation of episymbiont communities. Male Jaera are usually smaller than females despite evidence of sexual selection for body size and size-assortative mating [40, 94], suggesting that the effect of body size on microbiome composition would be worth investigating further [95]. Male Jaera could also potentially occupy different microhabitats than females as a consequence of sexual

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 12 / 19 Microbiome composition in Jaera albifrons spp.

dimorphism [96]. Although all samples were collected from underneath rocks, which always harbour mixed sex populations [46, 47], sex-specific differences in substrate microhabitat occupation cannot be ruled out but are yet to be investigated. Similarly, behavioural differences could also affect both epi- and endosymbiont communities, for example through differences in feeding rates or food preferences among sexes, which is well supported in copepods [97]. Gravid females in particular would be expected to change feeding habits or even cease feeding altogether, as is the case in the amphipod Ligia [98]. All sampled Jaera females were not gravid, but sexual receptivity or other ongoing reproductive processes may well have caused sex differences among microbiomes, particularly during the reproductive peak in summer where these differences were most pronounced [46].

Outlook and conclusions In summary, the Jaera microbiome is highly diverse and appears to be subject to multiple spatio-temporal environmental sources of variation, which is typical of marine intertidal microbiomes. A surprising result was the weak evidence of sex-ratio distorting reproductive parasites, which suggested very low infection levels at best in spite of pervasive sex-ratio distortion. How- ever, the finding of sex-specific patterns in overall microbiome composition warrants closer scrutiny and establishes the Jaera albifrons species complex as an intriguing study system for the effects of microbiomes on host reproductive processes. Our study has provided a snapshot assay that highlights the vast amount of variation in microbiomes from highly dynamic and complex environments. No doubt much of the ephemeral variation that has been characterised is attributed to ephemeral epibionts that may not necessarily be linked to host metabolism. Variation in this fraction could be reduced by maintaining Jaera long-term under controlled common-garden conditions [38]. Simi- larly, dissecting digestive or reproductive tracts could be a worthwhile avenue for targeting more specialised symbionts, since these tissues are often dominated by few taxa in tight association with host metabolism [73, 99–101]. Notwithstanding, our study highlights that, in order to properly understand the causes and consequences of phylosymbiosis or other effects of the hologenome it is essential to characterise microbiomes in situ with an appropriate sampling design that allows for appreciation of all sources of extrinsic and intrinsic variation. As such, more descriptive studies are essential for generating hypotheses of how the hologenome may operate in complex environments beyond classic model systems or controlled labo- ratory environments [34].

Supporting information S1 Fig. Rarefaction curves for male and female samples in single-end and paired-end assemblies. (PDF) S2 Fig. Microbial diversity metrics in samples grouped by season, region and sex. (PDF) S3 Fig. Median relative abundance of predicted KEGG pathways. (PDF) S4 Fig. Hierarchical clustering and multidimensional scaling of Jaccard and Bray-Curtis dissimilarities in single-end and paired-end assemblies. (PDF)

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 13 / 19 Microbiome composition in Jaera albifrons spp.

S5 Fig. Fold changes and taxonomic classification of sequence variants with differential abundance between sexes (northern region only), based on rarefied data. The top panels represent the full dataset of eleven samples; the bottom panels represent the eight samples from the northern region only. (PDF) S6 Fig. Fold changes and taxonomic classification of sequence variants with differential abundance between sexes, based on non-rarefied data. The top panels represent the full dataset of eleven samples; the bottom panels represent the eight samples from the northern region only. (PDF) S1 Table. Total read counts and relative abundances of sequences closely related to SILVA- strains of reproductive parasites Wolbachia, Rickettsia, Spiroplasma and Cardinium. (XLSX)

Acknowledgments We are grateful to Heather Ritchie and Laura Howell for fieldwork assistance. We acknowl- edge the computational support of the Maxwell HPC cluster funded by the University of Aberdeen.

Author Contributions Conceptualization: Alex Douglas, Stuart B. Piertney. Funding acquisition: Alex Douglas, Stuart B. Piertney. Investigation: Marius A. Wenzel. Methodology: Marius A. Wenzel. Project administration: Alex Douglas, Stuart B. Piertney. Writing – original draft: Marius A. Wenzel, Stuart B. Piertney. Writing – review & editing: Marius A. Wenzel, Alex Douglas, Stuart B. Piertney.

References 1. Holmes E, Li JV, Athanasiou T, Ashrafian H, Nicholson JK. Understanding the role of gut microbiome± host metabolic signal disruption in health and disease. Trends Microbiol. 2011; 19(7):349±359. https:// doi.org/10.1016/j.tim.2011.05.006 PMID: 21684749 2. Lyte M. Microbial endocrinology in the microbiome-gut-brain axis: how bacterial production and utilization of neurochemicals influence behavior. PLoS Pathog. 2013; 9(11):e1003726. https://doi.org/10. 1371/journal.ppat.1003726 PMID: 24244158 3. Young VB. The role of the microbiome in human health and disease: an introduction for clinicians. BMJ. 2017; 356. 4. Turner TR, James EK, Poole PS. The plant microbiome. Genome Biol. 2013; 14(6):209. https://doi. org/10.1186/gb-2013-14-6-209 PMID: 23805896 5. Fitzpatrick CR, Copeland J, Wang PW, Guttman DS, Kotanen PM, Johnson MT. Assembly and ecological function of the root microbiome across angiosperm plant species. Proc Natl Acad Sci USA. 2018; p. 201717617. 6. Mouchka ME, Hewson I, Harvell CD. Coral-associated bacterial assemblages: current knowledge and the potential for climate-driven impacts. Integr Comp Biol. 2010; 50(4):662±674. https://doi.org/10. 1093/icb/icq061 PMID: 21558231

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 14 / 19 Microbiome composition in Jaera albifrons spp.

7. Ainsworth TD, Krause L, Bridge T, Torda G, Raina JB, Zakrzewski M, et al. The coral core microbiome identifies rare bacterial taxa as ubiquitous endosymbionts. ISME J. 2015; 9(10):2261. https://doi.org/ 10.1038/ismej.2015.39 8. Roger AJ, Muñoz-GoÂmez SA, Kamikawa R. The Origin and Diversification of Mitochondria. Curr Biol. 2017; 27(21):R1177±R1192. https://doi.org/10.1016/j.cub.2017.09.015 PMID: 29112874 9. Nakayama T, Archibald JM. Evolving a photosynthetic organelle. BMC Biol. 2012; 10(1):35. https:// doi.org/10.1186/1741-7007-10-35 PMID: 22531210 10. Haddock SH, Moline MA, Case JF. Bioluminescence in the sea. Annu Rev Mar Sci. 2010; 2:443±493. https://doi.org/10.1146/annurev-marine-120308-081028 11. McFall-Ngai MJ. The importance of microbes in animal development: lessons from the squid-vibrio symbiosis. Annu Rev Microbiol. 2014; 68:177±194. https://doi.org/10.1146/annurev-micro-091313- 103654 PMID: 24995875 12. Sharon G, Segal D, Ringo JM, Hefetz A, Zilber-Rosenberg I, Rosenberg E. Commensal bacteria play a role in mating preference of Drosophila melanogaster. Proc Natl Acad Sci USA. 2010; 107 (46):20051±20056. https://doi.org/10.1073/pnas.1009906107 PMID: 21041648 13. Brucker RM, Bordenstein SR. The hologenomic basis of speciation: gut bacteria cause hybrid lethality in the genus Nasonia. Science. 2013; 341(6146):667±669. https://doi.org/10.1126/science.1240659 PMID: 23868918 14. Bordenstein SR, Theis KR. Host biology in light of the microbiome: ten principles of holobionts and hologenomes. PLoS Biol. 2015; 13(8):e1002226. https://doi.org/10.1371/journal.pbio.1002226 PMID: 26284777 15. Gilbert SF, Bosch TC, LedoÂn-Rettig C. Eco-Evo-Devo: developmental symbiosis and developmental plasticity as evolutionary agents. Nat Rev Genet. 2015; 16(10):611. https://doi.org/10.1038/nrg3982 PMID: 26370902 16. Shropshire JD, Bordenstein SR. Speciation by symbiosis: the microbiome and behavior. mBio. 2016; 7(2):e01785±15. https://doi.org/10.1128/mBio.01785-15 PMID: 27034284 17. Sunagawa S, Coelho LP, Chaffron S, Kultima JR, Labadie K, Salazar G, et al. Structure and function of the global ocean microbiome. Science. 2015; 348(6237):1261359. https://doi.org/10.1126/science. 1261359 PMID: 25999513 18. Rosenberg E, Zilber-Rosenberg I. Microbes drive evolution of animals and plants: the hologenome concept. mBio. 2016; 7(2):e01395±15. https://doi.org/10.1128/mBio.01395-15 PMID: 27034283 19. Charlat S, Hurst GD, MercËot H. Evolutionary consequences of Wolbachia infections. Trends Genet. 2003; 19(4):217±223. https://doi.org/10.1016/S0168-9525(03)00024-6 PMID: 12683975 20. Hagimori T, Abe Y, Date S, Miura K. The first finding of a Rickettsia bacterium associated with parthenogenesis induction among insects. Curr Microbiol. 2006; 52(2):97±101. https://doi.org/10.1007/ s00284-005-0092-0 PMID: 16450063 21. Duron O, Bouchon D, Boutin S, Bellamy L, Zhou L, EngelstaÈdter J, et al. The diversity of reproductive parasites among arthropods: Wolbachia do not walk alone. BMC Biol. 2008; 6(1):27. https://doi.org/ 10.1186/1741-7007-6-27 PMID: 18577218 22. Werren JH, Baldo L, Clark ME. Wolbachia: master manipulators of invertebrate biology. Nat Rev Microbiol. 2008; 6(10):741±751. https://doi.org/10.1038/nrmicro1969 PMID: 18794912 23. Smith J. The ecology and evolution of microsporidian parasites. Parasitology. 2009; 136(14):1901± 1914. https://doi.org/10.1017/S0031182009991818 PMID: 19995469 24. Kageyama D, Ohno M, Sasaki T, Yoshido A, Konagaya T, Jouraku A, et al. Feminizing Wolbachia endosymbiont disrupts maternal sex chromosome inheritance in a butterfly species. Evol Lett. 2017; 1 (5):232±244. https://doi.org/10.1002/evl3.28 25. Hurst GD, Jiggins FM. Problems with mitochondrial DNA as a marker in population, phylogeographic and phylogenetic studies: the effects of inherited symbionts. Proc Royal Soc Lond B. 2005; 272 (1572):1525±1534. https://doi.org/10.1098/rspb.2005.3056 26. JaÈckel R, Mora D, Dobler S. Evidence for selective sweeps by Wolbachia infections: phylogeny of Altica leaf beetles and their reproductive parasites. Mol Ecol. 2013; 22(16):4241±4255. https://doi.org/ 10.1111/mec.12389 PMID: 23927410 27. Jaenike J, Dyer KA, Cornish C, Minhas MS. Asymmetrical reinforcement and Wolbachia infection in Drosophila. PLoS Biol. 2006; 4(10):e325. https://doi.org/10.1371/journal.pbio.0040325 PMID: 17032063 28. Bordenstein SR, Werren JH. Bidirectional incompatibility among divergent Wolbachia and incompatibility level differences among closely related Wolbachia in Nasonia. Heredity. 2007; 99(3):278. https:// doi.org/10.1038/sj.hdy.6800994 PMID: 17519968

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 15 / 19 Microbiome composition in Jaera albifrons spp.

29. Jing X, Wong ACN, Chaston JM, Colvin J, McKenzie CL, Douglas AE. The bacterial communities in plant phloem-sap-feeding insects. Mol Ecol. 2014; 23(6):1433±1444. https://doi.org/10.1111/mec. 12637 PMID: 24350573 30. Dittmer J, Beltran-Bech S, Lesobre J, Raimond M, Johnson M, Bouchon D. Host tissues as microhabitats for Wolbachia and quantitative insights into the bacterial community in terrestrial isopods. Mol Ecol. 2014; 23(10):2619±2635. https://doi.org/10.1111/mec.12760 PMID: 24750488 31. Dittmer J, Lesobre J, Moumen B, Bouchon D. Host origin and tissue microhabitat shaping the microbiota of the terrestrial isopod Armadillidium vulgare. FEMS Microbiol Ecol. 2016; 92(5):fiw063. https:// doi.org/10.1093/femsec/fiw063 PMID: 27004796 32. Simhadri RK, Fast EM, Guo R, Schultz MJ, Vaisman N, Ortiz L, et al. The gut commensal microbiome of Drosophila melanogaster is modified by the endosymbiont Wolbachia. mSphere. 2017; 2(5): e00287±17. https://doi.org/10.1128/mSphere.00287-17 PMID: 28932814 33. Bahrndorff S, Alemu T, Alemneh T, Lund Nielsen J. The microbiome of animals: implications for con- servation biology. Int J Genomics. 2016; 2016. https://doi.org/10.1155/2016/5304028 PMID: 27195280 34. Hird SM. Evolutionary biology needs wild microbiomes. Front Microbiol. 2017; 8. https://doi.org/10. 3389/fmicb.2017.00725 35. Moisander PH, Sexton AD, Daley MC. Stable associations masked by temporal variability in the marine copepod microbiome. PLoS One. 2015; 10(9):e0138967. https://doi.org/10.1371/journal.pone. 0138967 PMID: 26393930 36. Weigel BL, Erwin PM. Intraspecific variation in microbial symbiont communities of the sun sponge, Hymeniacidon heliophila, from intertidal and subtidal habitats. Appl Environ Microbiol. 2016; 82 (2):650±658. https://doi.org/10.1128/AEM.02980-15 PMID: 26567307 37. Schuelke T, Pereira TJ, Hardy SM, Bik HM. Nematode-associated microbial taxa do not correlate with host phylogeny, geographic region or feeding morphology in marine sediment habitats. Mol Ecol. 2018;. https://doi.org/10.1111/mec.14539 PMID: 29600535 38. Brooks AW, Kohl KD, Brucker RM, van Opstal EJ, Bordenstein SR. Phylosymbiosis: relationships and functional effects of microbial communities across host evolutionary history. PLoS Biol. 2016; 14(11): e2000225. https://doi.org/10.1371/journal.pbio.2000225 PMID: 27861590 39. Bocquet C. Evolution of a superspecies of marine isopods. Syst Zool. 1954; 3(4):149±162. https://doi. org/10.2307/2411376 40. Naylor E, et al. British marine isopods. Academic Press for the Linnean Society of London; 1972. 41. Solignac M. Isolating mechanisms and modalities of speciation in the Jaera albifrons species complex (Crustacea, Isopoda). Syst Biol. 1981; 30(4):387±405. https://doi.org/10.1093/sysbio/30.4.387 42. Mifsud DV. The genetic basis of speciation in the Jaera albifrons species group of intertidal isopods [PhD thesis]. University of Aberdeen; 2011. Available from: http://ethos.bl.uk/OrderDetails.do?uin=uk. bl.ethos.540343. 43. Naylor E, Haahtela I. Habitat preferences and interspersion of species within the superspecies Jaera albifrons Leach (Crustacea: Isopoda). J Anim Ecol. 1966; p. 209±216. https://doi.org/10.2307/2697 44. Jones M. Effects of salinity on the survival of the Jaera albifrons Leach group of species (Crustacea: Isopoda). J Exp Mar Biol Ecol. 1972; 9(3):231±237. https://doi.org/10.1016/0022-0981(72)90035-4 45. Ribardière A, Daguin-ThieÂbaut C, Houbin C, Coudret J, Broudin C, Timsit O, et al. Geographically dis- tinct patterns of reproductive isolation and hybridization in two sympatric species of the Jaera albifrons complex (marine isopods). Ecol Evol. 2017; 7(14):5352±5365. https://doi.org/10.1002/ece3.3106 PMID: 28770073 46. Jones M, Naylor E. Breeding and bionomics of the British members of the Jaera albifrons group of species (Isopoda: Asellota). J Zool. 1971; 165(2):183±199. https://doi.org/10.1111/j.1469-7998.1971. tb02181.x 47. Piertney SB, Carvalho GR. Sex ratio variation in the intertidal isopod, Jaera albifrons. J Mar Biol Assoc UK. 1996; 76(03):825±828. https://doi.org/10.1017/S0025315400031490 48. Ribardière A, Centanni J, Dano A, Coudret J, Daguin-Thiebaut C, Houbin C, et al. Female-biased sex ratios unrelated to Wolbachia infection in European species of the Jaera albifrons complex (marine isopods). J Exp Mar Biol Ecol. 2018;submitted. 49. Bouchon D, Rigaud T, Juchault P. Evidence for widespread Wolbachia infection in isopod crustaceans: molecular identification and host feminization. Proc Royal Soc Lond B. 1998; 265(1401):1081± 1090. https://doi.org/10.1098/rspb.1998.0402 50. Stanhope MJ, Connelly MM, Hartwick B. Evolution of a crustacean chemical communication channel: behavioral and ecological genetic evidence for a habitat-modified, race-specific pheromone. J Chem Ecol. 1992; 18(10):1871±1887. https://doi.org/10.1007/BF02751110 PMID: 24254727

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 16 / 19 Microbiome composition in Jaera albifrons spp.

51. Klindworth A, Pruesse E, Schweer T, Peplies J, Quast C, Horn M, et al. Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 2013; 41(1):e1±e1. https://doi.org/10.1093/nar/gks808 PMID: 22933715 52. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods. 2016; 13(7):581±583. https://doi.org/10. 1038/nmeth.3869 PMID: 27214047 53. R Core Team. R: A Language and Environment for Statistical Computing; 2017. Available from: https://www.R-project.org/. 54. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2012; 41(D1): D590±D596. https://doi.org/10.1093/nar/gks1219 PMID: 23193283 55. Wang Q, Garrity GM, Tiedje JM, Cole JR. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol. 2007; 73(16):5261±5267. https:// doi.org/10.1128/AEM.00062-07 PMID: 17586664 56. Edgar RC. Updating the 97 bioRxiv. 2017; p. 192211. 57. Aûhauer KP, Wemheuer B, Daniel R, Meinicke P. Tax4Fun: predicting functional profiles from meta- genomic 16S rRNA data. Bioinformatics. 2015; 31(17):2882±2884. https://doi.org/10.1093/ bioinformatics/btv287 PMID: 25957349 58. Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 2015; 44(D1):D457±D462. https://doi.org/10.1093/nar/ gkv1070 PMID: 26476454 59. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in per- formance and usability. Mol Biol Evol. 2013; 30(4):772±780. https://doi.org/10.1093/molbev/mst010 PMID: 23329690 60. Paradis E, Claude J, Strimmer K. APE: analyses of phylogenetics and evolution in R language. Bioin- formatics. 2004; 20:289±290. https://doi.org/10.1093/bioinformatics/btg412 PMID: 14734327 61. Morgulis A, Coulouris G, Raytselis Y, Madden TL, Agarwala R, SchaÈffer AA. Database indexing for production MegaBLAST searches. Bioinformatics. 2008; 24(16):1757±1764. https://doi.org/10.1093/ bioinformatics/btn322 PMID: 18567917 62. McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013; 8(4):e61217. https://doi.org/10.1371/journal.pone. 0061217 PMID: 23630581 63. Venables WN, Ripley BD. Modern applied statistics with S. 4th ed. New York: Springer; 2002. Avail- able from: http://www.stats.ox.ac.uk/pub/MASS4. 64. Mariadassou M. Github repository; 2016. Available from: https://github.com/mahendra-mariadassou/ phyloseq-extended. 65. Yu G, Smith DK, Zhu H, Guan Y, Lam TTY. ggtree: an R package for visualization and annotation of phylogenetic trees with their covariates and other associated data. Methods Ecol Evol. 2017; 8(1):28± 36. https://doi.org/10.1111/2041-210X.12628 66. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. vegan: community ecology package; 2017. Available from: https://CRAN.R-project.org/package=vegan. 67. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15(12):550. https://doi.org/10.1186/s13059-014-0550-8 PMID: 25516281 68. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Royal Stat Soc Series B. 1995; 57:289±300. 69. Gilbert JA, Field D, Swift P, Newbold L, Oliver A, Smyth T, et al. The seasonal structure of microbial communities in the Western English Channel. Environ Microbiol. 2009; 11(12):3132±3139. https://doi. org/10.1111/j.1462-2920.2009.02017.x PMID: 19659500 70. Mengoni A, Focardi A, Bacci G, Ugolini A. High genetic diversity and variability of bacterial communities associated with the sandhopper Talitrus saltator (Montagu)(Crustacea, Amphipoda). Estuarine Coastal Shelf Sci. 2013; 131:75±82. https://doi.org/10.1016/j.ecss.2013.08.011 71. Bolhuis H, Cretoiu MS. What is so special about marine microorganisms? Introduction to the marine microbiomeÐfrom diversity to biotechnological potential. In: The Marine Microbiome. Springer; 2016. p. 3±20. 72. Degli Esposti M, Romero EM. The functional microbiome of arthropods. PLoS One. 2017; 12(5): e0176573. https://doi.org/10.1371/journal.pone.0176573 PMID: 28475624

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 17 / 19 Microbiome composition in Jaera albifrons spp.

73. Hakim JA, Koo H, Kumar R, Lefkowitz EJ, Morrow CD, Powell ML, et al. The gut microbiome of the sea urchin, Lytechinus variegatus, from its natural habitat demonstrates selective attributes of microbial taxa and predictive metabolic profiles. FEMS Microbiol Ecol. 2016; 92(9):fiw146. https://doi.org/ 10.1093/femsec/fiw146 PMID: 27368709 74. Carman KR, Dobbs FC. Epibiotic microorganisms on copepods and other marine crustaceans. Microsc Res Tech. 1997; 37(2):116±135. https://doi.org/10.1002/(SICI)1097-0029(19970415) 37:2<116::AID-JEMT2>3.0.CO;2-M PMID: 9145394 75. Wang XW, Wang JX. Crustacean hemolymph microbiota: Endemic, tightly controlled, and utilization expectable. Mol Immunol. 2015; 68(2):404±411. https://doi.org/10.1016/j.molimm.2015.06.018 PMID: 26153452 76. Pavloudi C, Kristoffersen JB, Oulas A, De Troch M, Arvanitidis C. Sediment microbial taxonomic and functional diversity in a natural salinity gradient challenge Remane's ªspecies minimumº concept. PeerJ. 2017; 5:e3687. https://doi.org/10.7717/peerj.3687 PMID: 29043106 77. Cordaux R, Pichon S, Hatira HBA, Doublet V, Grève P, MarcadeÂ I, et al. Widespread Wolbachia infection in terrestrial isopods and other crustaceans. Zookeys. 2012;(176):123. https://doi.org/10.3897/ zookeys.176.2284 PMID: 22536103 78. Baldini F, Segata N, Pompon J, Marcenac P, Shaw WR, DabireÂ RK, et al. Evidence of natural Wolba- chia infections in field populations of Anopheles gambiae. Nat Commun. 2014; 5:3985. https://doi.org/ 10.1038/ncomms4985 PMID: 24905191 79. Rogers EE, Backus EA. Anterior foregut microbiota of the glassy-winged sharpshooter explored using deep 16S rRNA gene sequencing from individual insects. PLoS One. 2014; 9(9):e106215. https://doi. org/10.1371/journal.pone.0106215 PMID: 25184624 80. Yong HS, Song SL, Chua KO, Lim PE. Predominance of Wolbachia endosymbiont in the microbiota across life stages of Bactrocera latifrons (Insecta: Tephritidae). Meta Gene. 2017; 14:6±11. https://doi. org/10.1016/j.mgene.2017.07.007 81. Mee PT, Weeks AR, Walker PJ, Hoffmann AA, Duchemin JB. Detection of low-level Cardinium and Wolbachia infections in Culicoides. Appl Environ Microbiol. 2015; 81(18):6177±6188. https://doi.org/ 10.1128/AEM.01239-15 PMID: 26150447 82. Larsen AM, Bullard SA, Womble M, Arias CR. Community structure of skin microbiome of gulf killifish, Fundulus grandis, is driven by seasonality and not exposure to oiled sediments in a Louisiana salt marsh. Microb Ecol. 2015; 70(2):534±544. https://doi.org/10.1007/s00248-015-0578-7 PMID: 25704317 83. Bierlich KC, Miller C, DeForce E, Friedlaender AS, Johnston DW, Apprill A. Temporal and regional variability in the skin microbiome of humpback whales along the Western Antarctic Peninsula. Appl Environ Microbiol. 2018; 84(5):e02574±17. https://doi.org/10.1128/AEM.02574-17 PMID: 29269499 84. Winters AD, Marsh TL, Brenden TO, Faisal M. Analysis of bacterial communities associated with the benthic amphipod Diporeia in the Laurentian Great Lakes Basin. Can J Microbiol. 2014; 61(1):72±81. https://doi.org/10.1139/cjm-2014-0434 85. Lokesh J, Kiron V. Transition from freshwater to seawater reshapes the skin-associated microbiota of Atlantic salmon. Sci Rep. 2016; 6. https://doi.org/10.1038/srep19707 PMID: 26806545 86. Hjerde E, Lorentzen MS, Holden MT, Seeger K, Paulsen S, Bason N, et al. The genome sequence of the fish pathogen Aliivibrio salmonicida strain LFI1238 shows extensive evidence of gene decay. BMC Genomics. 2008; 9(1):616. https://doi.org/10.1186/1471-2164-9-616 PMID: 19099551 87. Chistoserdov AY, Quinn RA, Gubbala SL, Smolowitz R. Bacterial communities associated with lesions of shell disease in the American lobster, Homarus americanus Milne-Edwards. J Shellfish Res. 2012; 31(2):449±462. https://doi.org/10.2983/035.031.0205 88. Kessner L, Spinard E, Gomez-Chiarri M, Rowley DC, Nelson DR. Draft genome sequence of Aliiro- seovarius crassostreae CV919-312, the causative agent of Roseovarius oyster disease (formerly juvenile oyster disease). Genome Announc. 2016; 4(2):e00148±16. https://doi.org/10.1128/genomeA. 00148-16 PMID: 26988054 89. Hazra S. Microbial Communities Associated with Marine Arthropods [PhD thesis]. University of Louisi- ana at Lafayette; 2014. Available from: https://search.proquest.com/docview/1669973383. 90. Gromek SM, Suria AM, Fullmer MS, Garcia JL, Gogarten JP, Nyholm SV, et al. Leisingera sp. JC1, a bacterial isolate from Hawaiian bobtail squid eggs, produces indigoidine and differentially inhibits vib- rios. Front Microbiol. 2016; 7. https://doi.org/10.3389/fmicb.2016.01342 PMID: 27660622 91. Martin MO, Gilman FR, Weiss SL. Sex-specific asymmetry within the cloacal microbiota of the striped plateau lizard, Sceloporus virgatus. Symbiosis. 2010; 51(1):97±105. https://doi.org/10.1007/s13199- 010-0078-y

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 18 / 19 Microbiome composition in Jaera albifrons spp.

92. Heuschele J, Selander E. The chemical ecology of copepods. J Plankton Res. 2014; 36(4):895±913. https://doi.org/10.1093/plankt/fbu025 93. Sutton R, Bolton E, Bartels-Hardege H, Eswards M, Reish D, Hardege J. Chemical signal mediated premating reproductive isolation in a marine polychaete, Neanthes acuminata (Arenaceodentata). J Chem Ecol. 2005; 31(8):1865±1876. https://doi.org/10.1007/s10886-005-5931-8 PMID: 16222812 94. Veuille M. Sexual behaviour and evolution of sexual dimorphism in body size in Jaera (Isopoda Asel- lota). Biol J Linn Soc. 1980; 13(1):89±100. https://doi.org/10.1111/j.1095-8312.1980.tb00072.x 95. Ishak HD, Miller JL, Sen R, Dowd SE, Meyer E, Mueller UG. Microbiomes of ant castes implicate new microbial roles in the fungus-growing ant Trachymyrmex septentrionalis. Sci Rep. 2011; 1:204. https:// doi.org/10.1038/srep00204 PMID: 22355719 96. Merilaita S, Jormalainen V. Evolution of sex differences in microhabitat choice and colour polymor- phism in Idotea baltica. Anim Behav. 1997; 54(4):769±778. https://doi.org/10.1006/anbe.1996.0490 PMID: 9344431 97. Saage A, Vadstein O, Sommer U. Feeding behaviour of adult Centropages hamatus (Copepoda, Calanoida): functional response and selective feeding experiments. J Sea Res. 2009; 62(1):16±21. https://doi.org/10.1016/j.seares.2009.01.002 98. Pennings SC, Carefoot TH, Zimmer M, Danko JP, Ziegler A. Feeding preferences of supralittoral isopods and amphipods. Can J Zool. 2000; 78(11):1918±1929. https://doi.org/10.1139/z00-143 99. Engel P, Moran NA. The gut microbiota of insects±diversity in structure and function. FEMS Microbiol Rev. 2013; 37(5):699±735. https://doi.org/10.1111/1574-6976.12025 PMID: 23692388 100. King GM, Judd C, Kuske CR, Smith C. Analysis of stomach and gut microbiomes of the eastern oyster (Crassostrea virginica) from coastal Louisiana, USA. PLoS One. 2012; 7(12):e51475. https://doi.org/ 10.1371/journal.pone.0051475 PMID: 23251548 101. Viver T, Orellana LH, Hatt JK, Urdiain M, DõÂaz S, Richter M, et al. The low diverse gastric microbiome of the jellyfish Cotylorhiza tuberculata is dominated by four novel taxa. Environ Microbiol. 2017; 19 (8):3039±3058. https://doi.org/10.1111/1462-2920.13763 PMID: 28419691

PLOS ONE | https://doi.org/10.1371/journal.pone.0202212 August 29, 2018 19 / 19